India cuts AI deepfake removal deadline from 36 hours to 3
India's amended IT Rules require platforms to remove unlawful AI-generated content within 3 hours of notice, down from 36, label all synthetic content, and proactively detect illegal material.
India has tightened the rules governing AI-generated content on social media platforms, cutting the deadline for removing unlawful synthetic material from 36 hours to 3, and mandating clear labelling of AI-generated content across the board.
What changed
Under amendments to the IT Rules made in February, platforms must label permissible AI-generated content and embed traceable metadata so users can identify synthetically generated material — audio, video, or text. That's the baseline for content that isn't itself illegal.
Unlawful AI-generated content — deepfakes, child sexual abuse material, non-consensual intimate images, AI-based impersonation — gets stricter treatment. Platforms must tell users about the legal consequences of creating or sharing this material, and once a valid notice arrives from the government or a court, they now have 3 hours to remove it. It was 36.
Platforms also have to deploy "reasonable and appropriate technical measures" to stop users creating or spreading unlawful synthetic content in the first place, not just respond after it's up.
Proactive detection is now required
Significant social media intermediaries — the larger platforms, in the IT Rules' terminology — must use automated tools to proactively detect specific categories of unlawful content, including material depicting rape or child sexual abuse. They also have to make "reasonable efforts" to catch content matching material already removed. In practice, that's a hash-matching or fingerprinting requirement.
Grievance timelines shrank too
Complaints involving sensitive categories like nudity and impersonation now need addressing within 2 hours, down from 24. That's a much tighter operational bar for platform trust and safety teams.
What's at stake for non-compliance
Failure to comply puts a platform's safe-harbour protection under Section 79 of the IT Act at risk — the provision that shields intermediaries from liability for content their users post, subject to conditions. Losing that protection is serious: it opens platforms to direct liability for user content they didn't create.
The framing
The government describes this as building an "open, safe, trusted and accountable cyberspace" while still promoting "responsible development, deployment and adoption of AI" in India. Set against the timeline compression — 36 hours to 3, 24 to 2 — the practical effect is a much faster enforcement clock for platforms operating in India, whatever the stated intent.